THE LAW OF JOHANNSEN. 221 



difference of their genotype from the set of genes of the 

 tail-less individual. 



We must see, that the similarity in size between a cell and 

 its daughter-cell, and the similarity in size and spininess of the 

 shell around those cells in uni-cellular organisms, and all 

 sameness of this nature is merely a sameness of the characters 

 which are for both cells the result of a cooperation of the same 

 genes, and the same non-genetic developmental characters. 

 There is no room for any great unlikeness through change in 

 the non-inherited developmental factors, through change in 

 anything but the genes. The similarity between a shell of 

 Difflugia and that of its "daughter" is not of the same order 

 as the similarity of two pea-plants of the same pure line, not 

 even of the same order as the similarity between two trees 

 budded from the same "mother-tree." And for the same reasons 

 the distinction between a clone of Difflugia with large shells 

 and one with small shells, is not of the same order, as the differ- 

 ence between a strain of mice with long tails and one with 

 short tails, and not even of the same order as the difference 

 between a clone of potatoes with large leaves and one with 

 small leaves. The difference between the clones of Difflugia 

 need not be due to genotypic difference at all, whereas the 

 dissimilarity between two clones of potatoes is almost certainly 

 due to a genotypic difference. 



If we take a dozen buds from one tree and graft them on 

 a dozen seedlings, and we observe that the twelve trees grow 

 to the same height, this similarity in height shows similarity 

 of genotype. If we split a certain tree in halves length-wise and 

 succeed in making both halves live, the similarity in height 

 is not of the same order as that, which we observed in our 

 clone of twelve budded trees. If we carefully split a big tree 

 in winter and count the number of leaves which each half pro- 

 duces in spring, we can calculate the coefficient of correlation 

 between these numbers as compared to the average number of 

 leaves on this sort of tree. This would be a piece of work di- 

 rectly comparable with the studies on Difflugia of Jennings 



